PIPELINES_ERRORS

Warning

SingleStore 9.0 gives you the opportunity to preview, evaluate, and provide feedback on new and upcoming features prior to their general availability. In the interim, SingleStore 8.9 is recommended for production workloads, which can later be upgraded to SingleStore 9.0.

This view contains exhaustive node-level error information that may originate at either a leaf node or aggregator node. If a row appears to be a duplicate, first verify that the row’s ERROR_MESSAGE entry does not start with LEAF ERROR (<host>:<port>). Error event information may be propagated from a leaf to an aggregator or from a leaf to another leaf. This text indicates the specific leaf node origin for the error.

The following pipeline engine variables are used to control various aspects of pipeline behavior. For more detailed information consult the Sync Variables Lists section of the Engine Variable List.

pipelines_errors_retention_minutes - controls the amount of time in minutes that a pipeline error is stored on disk.
ingest_errors_max_disk_space_mb - controls the maximum amount of disk space (MB) that is used to log errors for pipelines.

information_schema.PIPELINES_ERRORS Schema

Column Name	Description
`DATABASE_NAME`	The name of the database associated with the error.
`PIPELINE_NAME`	The name of the pipeline associated with the error.
`ERROR_UNIX_TIMESTAMP`	The time of the error event in Unix timestamp format.
`ERROR_TYPE`	Specifies what type of error occurred. Possible values are `Error` or `Warning`. `Error`: An error occurred, which may have stopped the pipeline if the `pipelines_stop_on_error` variable is set to `ON`. `Warning`: A warning occurred, which does not stop a pipeline.
`ERROR_CODE`	The error code for the error.
`ERROR_MESSAGE`	The message associated with the error. This value contains contextual information about the error that can be used for debugging purposes, including a stack trace if the error was caused by a transform failure.
`ERROR_KIND`	Specifies whether the error event was caused by internal or external factors. Possible values are `Internal`, `Extract`, `Transform`, or `Load`. `Internal`: An internal error occurred within the Pipelines feature itself. `Extract`: An error occurred during extraction from the data source. Extraction errors are typically caused by network availability or problems with the data source partitions. `Transform`: An error occurred during data transformation. The transform executable is typically the cause of transformation errors. `Load`: An error occurred when attempting to load data into the destination table. Load errors are typically caused by malformed CSV data or attempting to write invalid schema types into a column, such as a `NULL` value into a non-nullable column.
`STD_ERROR`	The text that may have been outputted during the data transformation. The origin of this text is the transform executable itself. This value can be empty or `NULL` if the `ERROR_KIND` value is not `Transform` or if no standard error text was outputted by the transform during failure.
`LOAD_DATA_LINE`	The text of a `LOAD DATA` statement that caused a parsing error while attempting to load data into the destination table. This value contains the invalid line of CSV text that failed to parse. Load errors are typically caused by malformed CSV data or attempting to write invalid schema types into a column, such as a `NULL` value into a non-nullable column. This value can be empty or `NULL` if the error didn’t occur during the loading phase of pipeline execution.
`LOAD_DATA_LINE_NUMBER`	The line number of a `LOAD DATA` statement that caused a parsing error while attempting to load data into the destination table. A `LOAD DATA` statement may consist of many lines, and this value indicates the specific invalid line. This line number can be correlated with the value of `LOAD_DATA_LINE`, which contains the invalid line’s text. This value may be empty or `NULL` if a `LOAD DATA` statement wasn’t associated with the error event.
`BATCH_ID`	The internal unique identifier for the batch that experienced an error event.
`ERROR_ID`	The internal unique identifier for the error event.
`BATCH_SOURCE_PARTITION_ID`	The data source’s partition ID for the batch. This value may be `NULL` if the error occurred on the master aggregator.
`BATCH_EARLIEST_OFFSET`	Specifies the earliest offset for the batch. This value indicates the start of the offset range for a batch, while `BATCH_LATEST_OFFSET` indicates the end of the offset range. This value may be `NULL` if the error occurred on the master aggregator.
`BATCH_LATEST_OFFSET`	Specifies the latest offset for the batch. This value indicates the end of the offset range for a batch, while `BATCH_EARLIEST_OFFSET` indicates the start of the offset range. This value may be `NULL` if the error occurred on the master aggregator.
`HOST`	The hostname or host IP address for the leaf node that processed the batch. This value may be `NULL` if `BATCH_SOURCE_PARTITION_ID` or `PARTITION` is also NULL. The combination of a batch’s `HOST`, `PORT`, and `PARTITION` identify the specific leaf node partition that attempted to load batch data.
`PORT`	The port number for the leaf node that processed the batch. This value may be `NULL` if `BATCH_SOURCE_PARTITION_ID` or `PARTITION` is also `NULL`. The combination of a batch’s `HOST`, `PORT`, and `PARTITION` identify the specific leaf node partition that attempted to load batch data.
`PARTITION`	Specifies the partition ID on a leaf node that processed the batch. This value may be `NULL` if `BATCH_SOURCE_PARTITION_ID` is also `NULL`. The combination of a batch’s `HOST`, `PORT`, and `PARTITION` identify the specific leaf node partition that attempted to load batch data.

PIPELINES_ERRORS

On this page

information_schema.PIPELINES_ERRORS Schema

Was this article helpful?

On this page

Was this article helpful?